pith. sign in

arxiv: 2605.21899 · v1 · pith:CDZMQJL3new · submitted 2026-05-21 · 📊 stat.CO · math.PR· math.ST· stat.TH

Mad Props: Parallelism in Markov Chain Monte Carlo Through the Lens of the Infinite Proposal Limit

Nathan E. Glatt-Holtz , Andrew J. Holbrook , Justin A. Krometis , Cecilia F. Mondaini This is my paper

Pith reviewed 2026-05-22 02:50 UTC · model grok-4.3

classification 📊 stat.CO math.PRmath.STstat.TH
keywords multiproposal MCMCMarkov chain Monte Carloinfinite proposal limitinvolutive kernelsparallel samplinglarge p regimestate space methods
0
0 comments X

The pith

The infinite proposal limit simplifies multiproposal MCMC kernels to reveal new algorithms and relationships via involutive theory.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper studies multiproposal Markov Chain Monte Carlo methods in the regime where the number of proposals p becomes very large. It identifies and examines several new algorithms that perform well in this limit, rules out some prior approaches, and uncovers connections between different MP-MCMC families. The work rests on a general theory of multiproposal involutive kernels applied to the limiting behavior of various proposal and acceptance mechanisms. These results matter for designing parallel samplers that better handle complex target distributions when computational resources allow many proposals per step.

Core claim

In the large p-limit, MP-MCMC transition kernels reduce to explicit forms that identify promising new methods such as Algorithms 1.1, 3.3 and 3.4, eliminate other extant schemes, and establish new equivalences and inclusions among multiproposal methodologies, all derived within the authors' general state space multiproposal involutive framework.

What carries the argument

Large p-limit kernels for MP-MCMC algorithms across classes of proposal and acceptance structures, obtained from the general state space multiproposal involutive theory.

If this is right

  • Certain new algorithms become optimal or near-optimal samplers once p is taken large.
  • Some previously proposed MP-MCMC variants are provably dominated or redundant in the infinite-proposal regime.
  • New inclusion and equivalence relations let practitioners combine or substitute methods without loss of limiting performance.
  • Parallel implementations can safely increase proposal counts to approach the derived limiting kernels for efficiency gains.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Finite but large p implementations could be tuned to approximate these limiting kernels for practical speed-ups on high-dimensional targets.
  • The limit analysis may connect multiproposal methods to ensemble or multiple-chain techniques that also exploit parallelism.
  • Direct comparison of mixing times between the new algorithms and classical single-proposal MCMC on benchmark problems would test the practical payoff.

Load-bearing premise

The validity and applicability of the general state space multiproposal involutive theory that the authors recently constructed.

What would settle it

A numerical simulation or explicit calculation showing that the predicted large-p kernel for one of the new algorithms produces incorrect acceptance probabilities or fails to match observed chain behavior on a concrete target distribution.

Figures

Figures reproduced from arXiv: 2605.21899 by Andrew J. Holbrook, Cecilia F. Mondaini, Justin A. Krometis, Nathan E. Glatt-Holtz.

Figure 1
Figure 1. Figure 1: As the number of proposals p grows large, a single iteration of the slingshot kernel converges to the target distribution. For each proposal count setting, we generate 10,000 independent samples from the slingshot kernel with Gaussian proposal distributions (a) N(4, 1 2 ) and (b) N(4, σ⋆ f 2 ) for a standard normal target. For faster convergence in p, we use (6.6) to obtain the optimal proposal standard de… view at source ↗
Figure 2
Figure 2. Figure 2: Left: Density of the tilted banana distribution given in ( [PITH_FULL_IMAGE:figures/full_fig_p045_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Approximation of the tilted banana distribution by the slingshot kernel, by number of proposals [PITH_FULL_IMAGE:figures/full_fig_p045_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Tuning σf for the tilted banana distribution, by number of proposals p. Each point represents a chain of 200,000 MCMC samples with σf adaptively tuned to a given target acceptance rate. Left: The choice of σf by target acceptance rate. Middle: The acceptance rate by target acceptance rate, showing that the target acceptance rate was achieved up to the limits of the Barker formulation. Right: Loss by accept… view at source ↗
Figure 5
Figure 5. Figure 5: Thanks to its simplified proposal and acceptance structures, the slingshot allows for easy off-the [PITH_FULL_IMAGE:figures/full_fig_p048_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: As the number of proposals grows moderately large, MCMC samples obtained using the slingshot [PITH_FULL_IMAGE:figures/full_fig_p048_6.png] view at source ↗
read the original abstract

Multiproposal MCMC (MP-MCMC) algorithms use clouds of proposals to efficiently traverse state spaces and overcome complex target geometries. While MCMC methods are embarrassingly parallel by nature, the non-trivial forms of parallelism provided by the MP-MCMC formalism sometimes leads to significant improvements over a naive approach. Here, one important tuning parameter is the number of proposals p used by a single MP-MCMC iteration. While a number of computational strategies have been proposed to efficiently leverage large numbers of proposals within the MP-MCMC paradigm, much remains unknown about these algorithms, particularly in the large p-regime. In this contribution, we discover surprising results by identifying and studying several promising new methods (Algorithm 1.1, Algorithm 3.3, Algorithm 3.4), ruling out other extant approaches and discovering new relationships between different MP-MCMC methodologies. Our analysis is centered on a general state space multiproposal involutive theory recently constructed by the authors combined with the consideration of the large p-limit kernels for MP-MCMC algorithms within a variety of different classes of proposal and acceptance structures.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper analyzes multiproposal MCMC (MP-MCMC) algorithms by taking the infinite-proposal limit (large p) of their transition kernels. Centered on the authors' recently developed general state-space multiproposal involutive theory, it identifies and studies new methods (Algorithms 1.1, 3.3, 3.4), rules out certain existing approaches, and uncovers relationships among MP-MCMC methodologies for different classes of proposal and acceptance structures.

Significance. If the limiting kernels are rigorously derived and remain well-defined without hidden regularity violations as p→∞, the work could clarify the asymptotic behavior of parallel MCMC proposals and guide selection among competing MP-MCMC strategies. Explicit identification of promising algorithms and falsifiable relationships between methods would be a concrete contribution to the theoretical understanding of parallelism in MCMC.

major comments (2)
  1. [Abstract and §1] Abstract and §1: The central claims about new algorithms, ruled-out approaches, and discovered relationships all rest on the validity of the authors' prior general state-space multiproposal involutive theory. The manuscript must supply explicit verification that the acceptance probabilities and transition kernels remain non-degenerate and correctly normalized in the p→∞ limit for the specific structures considered in Algorithms 1.1, 3.3, and 3.4; otherwise the performance rankings and inter-method relationships cannot be substantiated.
  2. [§3] §3 (or wherever the large-p kernels are derived): The skeptic note highlights that any unstated regularity condition on target or proposal densities violated in the infinite-proposal regime would invalidate the limiting kernels. The paper should include a concrete check or counter-example showing that the chosen proposal/acceptance structures satisfy the conditions needed for the limit to exist and match the claimed behavior.
minor comments (2)
  1. [Algorithm sections] Ensure that all algorithm pseudocode (1.1, 3.3, 3.4) includes explicit statements of the acceptance probability in the large-p limit rather than referring only to the general theory.
  2. [Throughout] Clarify notation for the number of proposals p versus other parameters to avoid confusion when discussing the p→∞ regime.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments, which have prompted us to strengthen the explicit verification of the limiting behavior in our analysis. We address each major comment below and have revised the manuscript to incorporate the requested checks and clarifications.

read point-by-point responses

  1. Referee: [Abstract and §1] Abstract and §1: The central claims about new algorithms, ruled-out approaches, and discovered relationships all rest on the validity of the authors' prior general state-space multiproposal involutive theory. The manuscript must supply explicit verification that the acceptance probabilities and transition kernels remain non-degenerate and correctly normalized in the p→∞ limit for the specific structures considered in Algorithms 1.1, 3.3, and 3.4; otherwise the performance rankings and inter-method relationships cannot be substantiated.

    Authors: We appreciate this observation and agree that explicit verification bolsters the claims. In the revised manuscript, we have added a dedicated subsection following the derivation of the large-p kernels. This subsection directly computes the limiting acceptance probabilities for Algorithms 1.1, 3.3, and 3.4 by applying the general involutive formula and taking the p→∞ limit term by term. We verify that each limiting acceptance probability lies strictly between 0 and 1 for non-degenerate cases, that the kernels integrate to 1, and that no hidden singularities arise under the proposal structures considered. These calculations confirm that the reported performance rankings and inter-method relationships are valid in the infinite-proposal regime. revision: yes

  2. Referee: [§3] §3 (or wherever the large-p kernels are derived): The skeptic note highlights that any unstated regularity condition on target or proposal densities violated in the infinite-proposal regime would invalidate the limiting kernels. The paper should include a concrete check or counter-example showing that the chosen proposal/acceptance structures satisfy the conditions needed for the limit to exist and match the claimed behavior.

    Authors: We concur that a concrete check is necessary. The revised Section 3 now includes an explicit verification step for each algorithm: we confirm that the integrability conditions and positivity requirements from the general multiproposal involutive theory hold uniformly in the p→∞ limit for the chosen proposal and acceptance structures. Direct computation shows that the relevant integrals converge to finite, positive values matching the claimed limiting kernels. We also briefly note why certain alternative structures (which we rule out) fail these checks, providing the requested counter-example context without invalidating the methods presented. revision: yes

Circularity Check

1 steps flagged

Central MP-MCMC large-p analysis and new relationships rest on authors' self-constructed involutive theory

specific steps
  1. self citation load bearing [Abstract]
    "Our analysis is centered on a general state space multiproposal involutive theory recently constructed by the authors combined with the consideration of the large p-limit kernels for MP-MCMC algorithms within a variety of different classes of proposal and acceptance structures."

    The strongest claims—identifying Algorithms 1.1, 3.3, 3.4, ruling out other approaches, and discovering new relationships—are obtained by taking large-p limits of kernels whose acceptance probabilities and well-definedness are furnished by the authors' own prior involutive theory. The validity of those limiting objects therefore depends on the correctness of the self-cited construction rather than on an independent mathematical fact or external check.

full rationale

The paper explicitly centers its derivation of new algorithms, performance rankings, and inter-method relationships on the large-p limits of kernels defined inside a general state-space multiproposal involutive framework that the same authors recently constructed. This matches the self-citation load-bearing pattern because the acceptance probabilities, transition kernels, and limiting objects used to obtain the claimed results are supplied by that prior construction; without independent verification or external benchmarks for the framework's behavior in the infinite-proposal regime, the strongest claims reduce to consequences of the self-cited theory. The analysis still contains new limiting calculations, so the circularity is partial rather than total.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the authors' recently constructed multiproposal involutive theory as the primary domain assumption; no free parameters or invented entities are identifiable from the abstract.

axioms (1)
  • domain assumption General state space multiproposal involutive theory recently constructed by the authors
    The analysis is centered on this theory combined with large p-limit kernels.

pith-pipeline@v0.9.0 · 5745 in / 1245 out tokens · 47744 ms · 2026-05-22T02:50:56.090562+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

116 extracted references · 116 canonical work pages · 1 internal anchor

  1. [1]

    and Lemey, P

    Holbrook, A.J. and Lemey, P. and Baele, G. and Dellicour, S. and Brockmann, D. and Rambaut, A. and Suchard, M.A. , title =. Journal of Computational and Graphical Statistics , volume =. 2021 , publisher =

    work page 2021

  2. [2]

    arXiv preprint arXiv:2304.06251 , year=

    Importance is Important: Generalized Markov Chain Importance Sampling Methods , author=. arXiv preprint arXiv:2304.06251 , year=

    work page arXiv

  3. [3]

    Electronic Journal of Statistics , volume=

    On a Metropolis--Hastings importance sampling estimator , author=. Electronic Journal of Statistics , volume=

    work page

  4. [4]

    Journal of Computational and Graphical Statistics , volume=

    Markov chain importance sampling—a highly efficient estimator for MCMC , author=. Journal of Computational and Graphical Statistics , volume=. 2020 , publisher=

    work page 2020

  5. [5]

    TensorFlow Distributions

    Tensorflow distributions , author=. arXiv preprint arXiv:1711.10604 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  6. [6]

    and Barham, P

    Abadi, M. and Barham, P. and Chen, J. and Chen, Z. and Davis, A. and Dean, J. and Devin, M. and Ghemawat, S. and Irving, G. and Isard, M. and others , booktitle=

    work page

  7. [7]

    , journal=

    Brockwell, A.E. , journal=. 2006 , publisher=

    work page 2006

  8. [8]

    and Loeffler, C.E

    Holbrook, A.J. and Loeffler, C.E. and Flaxman, S.R. and Suchard, M.A. , journal=. 2021 , publisher=

    work page 2021

  9. [9]

    and Ji, X

    Holbrook, A.J. and Ji, X. and Suchard, M.A. , journal=. 2022 , publisher=

    work page 2022

  10. [10]

    , journal=

    Green, P.J. , journal=. 1995 , publisher=

    work page 1995

  11. [11]

    , journal=

    Zanella, G. , journal=. 2020 , publisher=

    work page 2020

  12. [12]

    and Kennedy, A.D

    Duane, S. and Kennedy, A.D. and Pendleton, B.J. and Roweth, D. , journal=. 1987 , publisher=

    work page 1987

  13. [13]

    , journal=

    Rosenthal, J.S. , journal=. 2011 , publisher=

    work page 2011

  14. [14]

    and Calderhead, B

    Schwedes, T. and Calderhead, B. , booktitle=. 2021 , organization=

    work page 2021

  15. [15]

    and Calderhead, B

    Girolami, M. and Calderhead, B. , journal=. 2011 , publisher=

    work page 2011

  16. [16]

    Statistics and computing , volume=

    Sampling from multimodal distributions using tempered transitions , author=. Statistics and computing , volume=. 1996 , publisher=

    work page 1996

  17. [17]

    and Pillai, N.S

    Mangoubi, O. and Pillai, N.S. and Smith, A. , journal=

    work page

  18. [18]

    , journal=

    Neal, R.M. , journal=

    work page

  19. [19]

    Journal of multivariate analysis , volume=

    Generating MCMC proposals by randomly rotating the regular simplex , author=. Journal of multivariate analysis , volume=. 2023 , publisher=

    work page 2023

  20. [20]

    Transactions of Mathematics and its Applications , volume=

    Parallel MCMC algorithms: Theoretical foundations, algorithm design, case studies , author=. Transactions of Mathematics and its Applications , volume=

    work page

  21. [21]

    Quantum Speedups for Multiproposal

    Lin, Chin-Yi and Chen, Kuo-Chin and Lemey, Philippe and Suchard, Marc A and Holbrook, Andrew J and Hsieh, Min-Hsiu , journal=. Quantum Speedups for Multiproposal

    work page

  22. [22]

    Journal of Computational and Graphical Statistics , volume=

    A quantum parallel Markov chain Monte Carlo , author=. Journal of Computational and Graphical Statistics , volume=. 2023 , publisher=

    work page 2023

  23. [23]

    and Saksman, E

    Haario, H. and Saksman, E. and Tamminen, J. , journal=. 2001 , publisher=

    work page 2001

  24. [24]

    , year =

    Borchers, H.W. , year =

    work page

  25. [25]

    and Gilks, W.R

    Gelman, A. and Gilks, W.R. and Roberts, G.O. , journal=. 1997 , publisher=

    work page 1997

  26. [26]

    and Barber, D

    Williams, C.K. and Barber, D. , journal=. 1998 , publisher=

    work page 1998

  27. [27]

    , journal=

    Neal, R.M. , journal=. 2003 , publisher=

    work page 2003

  28. [28]

    and Gentleman, R

    Ihaka, R. and Gentleman, R. , journal=. 1996 , publisher=

    work page 1996

  29. [29]

    and Moraes, M.A

    Salmon, J.K. and Moraes, M.A. and Dror, R.O. and Shaw, D.E. , booktitle=

    work page

  30. [30]

    Statistical science , volume=

    Inference from iterative simulation using multiple sequences , author=. Statistical science , volume=. 1992 , publisher=

    work page 1992

  31. [31]

    Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume=

    Optimal scaling of discrete approximations to Langevin diffusions , author=. Journal of the Royal Statistical Society: Series B (Statistical Methodology) , volume=. 1998 , publisher=

    work page 1998

  32. [32]

    and Stramer, O

    Roberts, G.O. and Stramer, O. , journal=. 2002 , publisher=

    work page 2002

  33. [33]

    , journal=

    Scott, S.L. , journal=. 2002 , publisher=

    work page 2002

  34. [34]

    and Jourdain, B

    Delmas, J.-F. and Jourdain, B. , journal=. 2009 , publisher=

    work page 2009

  35. [35]

    , journal=

    Frenkel, D. , journal=. 2004 , publisher=

    work page 2004

  36. [36]

    and Tjelmeland, H

    Luo, X. and Tjelmeland, H. , journal=. 2019 , publisher=

    work page 2019

  37. [37]

    and Chen, Y

    Yang, S. and Chen, Y. and Bernton, E. and Liu, J.S. , journal=. 2018 , publisher=

    work page 2018

  38. [38]

    and Roberts, G.O

    Smith, A.F.M. and Roberts, G.O. , journal=. 1993 , publisher=

    work page 1993

  39. [39]

    and Welling, M

    Neklyudov, K. and Welling, M. and Egorov, E. and Vetrov, D. , booktitle=. 2020 , organization=

    work page 2020

  40. [40]

    The Annals of Applied Probability , number =

    Nathan Glatt-Holtz and Justin Krometis and Cecilia Mondaini , title =. The Annals of Applied Probability , number =. 2023 , doi =

    work page 2023

  41. [41]

    and Lee, A

    Andrieu, C. and Lee, A. and Livingstone, S. , journal=

    work page

  42. [42]

    Hastings, W. K. , title = ". Biometrika , volume =. 1970 , month =. doi:10.1093/biomet/57.1.97 , url =

    work page doi:10.1093/biomet/57.1.97 1970

  43. [43]

    The journal of chemical physics , volume=

    Equation of state calculations by fast computing machines , author=. The journal of chemical physics , volume=. 1953 , publisher=

    work page 1953

  44. [44]

    and Liang, F

    Liu, J.S. and Liang, F. and Wong, W.H. , journal=. 2000 , publisher=

    work page 2000

  45. [45]

    and Gelman, A

    Hoffman, M.D. and Gelman, A. , journal=

    work page

  46. [46]

    , journal=

    Barker, A.A. , journal=. 1965 , publisher=

    work page 1965

  47. [47]

    , journal=

    Calderhead, B. , journal=. 2014 , publisher=

    work page 2014

  48. [48]

    the Annals of Statistics , pages=

    Markov chains for exploring posterior distributions , author=. the Annals of Statistics , pages=. 1994 , publisher=

    work page 1994

  49. [49]

    SIAM Journal on Numerical Analysis , volume=

    The efficient generation of random orthogonal matrices with an application to condition estimators , author=. SIAM Journal on Numerical Analysis , volume=. 1980 , publisher=

    work page 1980

  50. [50]

    , journal=

    Tierney, L. , journal=. 1998 , publisher=

    work page 1998

  51. [51]

    and Roberts, G

    Beskos, A. and Roberts, G. and Stuart, A. and Voss, J. , journal=. 2008 , publisher=

    work page 2008

  52. [52]

    and Roberts, G.O

    Cotter, S.L. and Roberts, G.O. and Stuart, A.M. and White, D. , journal=. 2013 , publisher=

    work page 2013

  53. [53]

    and Stuart, A.M

    Dashti, M. and Stuart, A.M. , booktitle=. 2017 , publisher=

    work page 2017

  54. [54]

    , journal=

    Stuart, A.M. , journal=. 2010 , publisher=

    work page 2010

  55. [55]

    and Mondaini, C.F

    Glatt-Holtz, N.E. and Mondaini, C.F. , journal=. 2021 , publisher=

    work page 2021

  56. [56]

    and Simpson, D

    Terenin, A. and Simpson, D. and Draper, D. , booktitle=. 2020 , organization=

    work page 2020

  57. [57]

    2009 , publisher=

    Optimal transport: old and new , author=. 2009 , publisher=

    work page 2009

  58. [58]

    2014 , publisher=

    Stochastic equations in infinite dimensions , author=. 2014 , publisher=

    work page 2014

  59. [59]

    2006 , publisher=

    Statistical and computational inverse problems , author=. 2006 , publisher=

    work page 2006

  60. [60]

    and Glatt-Holtz, N

    Borggaard, J. and Glatt-Holtz, N. and Krometis, J. , journal=. 2020 , publisher=

    work page 2020

  61. [61]

    and Pinski, F.J

    Beskos, A. and Pinski, F.J. and Sanz-Serna, J.M. and Stuart, A.M. , journal=. 2011 , publisher=

    work page 2011

  62. [62]

    The Physics of Fluids , volume=

    Small-scale structure of a scalar field convected by turbulence , author=. The Physics of Fluids , volume=. 1968 , publisher=

    work page 1968

  63. [63]

    New Perspectives in Turbulence , pages=

    Stochastic modeling of isotropic turbulence , author=. New Perspectives in Turbulence , pages=. 1991 , publisher=

    work page 1991

  64. [64]

    and Carlin, J.B

    Gelman, A. and Carlin, J.B. and Stern, H.S. and Dunson, D.B. and Vehtari, A. and Rubin, D.B. , year =

    work page

  65. [65]

    2009 , journal =

    Gmsh: A three-dimensional finite element mesh generator with built-in pre-and post-processing pacilities , author =. 2009 , journal =

    work page 2009

  66. [66]

    Inverse Problems , author =

    A statistical framework for domain shape estimation in. Inverse Problems , author =. 2023 , note =. doi:10.1088/1361-6420/acdd8e , abstract =

    work page doi:10.1088/1361-6420/acdd8e 2023

  67. [67]

    , publisher =

    Bogachev, V.I. , publisher =. Measure theory , volume =

    work page

  68. [68]

    , publisher =

    Folland, G.B. , publisher =. Real analysis: modern techniques and their applications , volume =

    work page

  69. [69]

    2013 , publisher=

    Measure theory , author=. 2013 , publisher=

    work page 2013

  70. [70]

    1998 , publisher=

    Gaussian measures , author=. 1998 , publisher=

    work page 1998

  71. [71]

    and Border, K.C

    Aliprantis, C.D. and Border, K.C. , publisher =

    work page

  72. [72]

    2006 , publisher=

    Geometric numerical integration: structure-preserving algorithms for ordinary differential equations , author=. 2006 , publisher=

    work page 2006

  73. [73]

    and Reich, S

    Leimkuhler, B. and Reich, S. , number=. 2004 , publisher=

    work page 2004

  74. [74]

    and Sanz-Serna, J.M

    Bou-Rabee, N. and Sanz-Serna, J.M. , journal=. Geometric Integrators and the

    work page

  75. [75]

    and Stuart, A.M

    Hairer, M. and Stuart, A.M. and Vollmer, S.J. , journal=. 2014 , publisher=

    work page 2014

  76. [76]

    and Casella, G

    Robert, C.P. and Casella, G. , volume=. 1999 , publisher=

    work page 1999

  77. [77]

    and Robert, C.P

    Casella, G. and Robert, C.P. , journal=. 1996 , publisher=

    work page 1996

  78. [78]

    Communications in Mathematical Sciences , author =

    Conditional. Communications in Mathematical Sciences , author =. 2004 , note =

    work page 2004

  79. [79]

    , journal=

    Peskun, P.H. , journal=. 1973 , publisher=

    work page 1973

  80. [80]

    1997 , edition=

    A course in mathematical statistics , author=. 1997 , edition=

    work page 1997

Showing first 80 references.